TW 0 .05 . Using the Bonferroni inequality formula , the experimentwise Type I error rate can be computed as follows : As observed above , using five perfectly uncorrelated hypotheses tested at 5 percent per-comparison alpha (or testwise alpha , denoted (TW , the experimentwise alpha (denoted (EW ) was 0 . or roughly 22 .62 . This means that the probability of rejecting any or more than one of the hypotheses (or dependent variables ) is 22 .62 . However , determining the rate of experimentwise error does not inform...











SUBJECTS

